Jump to content
voxman55

Pod Go DSP processing power- has anyone yet hit a DSP limit?

Recommended Posts

I've picked up from watching videos of Line 6 Helix products (inc eg Helix Stomp & even Helix/Helix LT) that depending on the amp, cab & effects selected you can run out of processing power and it will 'grey out' certain options once the DSP processing power reaches its limit.  That makes perfect sense where you have very wide freedom to build your patches.

 

With Pod Go, I believe there were three effects Line 6 didn't include specifically because these were particularly demanding on DSP power.  Am I correct in my understanding that because Pod Go is already purpose designed to be 'self-limiting' from the perspective of DSP processing i.e. it has a 'fixed' series of blocks including only a single amp & cab option, plus 4 user assignable blocks plus an FX insert, that 'DSP overflow' is not an issue with Pod Go and that you can use it to its full capacity with any combination within the aforementioned framework  with no risk of DSP 'greying out' limitations regardless of what amp, cab, effects, insert you select?   

Share this post


Link to post
Share on other sites

No, you are not correct in your understanding. The POD Go uses dynamic DSP allocation as does Helix. Here is a snip from the POD Go manual where adding blocks to the signal path is discussed:
 

IMPORTANT! If you encounter items in the list that are grayed out or unavailable, this means there isn't enough DSP to accommodate that category or model. For example, if you've already added three reverbs, you probably won't be able to add a fourth.

Share this post


Link to post
Share on other sites

Thanks @silverhead. I'm certainly not intending to use 3 reverbs. But with only 4 assignable blocks I'd certainly want to have reasonable flexibility to use all the fixed blocks and all 4 assignable blocks, e.g. an amp, EQ, Volume/Wah, a cab, a distortion, a modulation, a reverb and or/a delay or possibly two delays e.g. one short and one long in a patch, and an FX loop insert stomp neffect.  Could I still be limited and cut down to e.g. only 3 assignable blocks? 

 

If so I'd like to more clearly understand what combinations or models are particularly DSP hungry and which combinations could cause problems. If this is likely to be a very rare occurrence by choosing three of something that's one thing, but if it's going to trigger more regular issues on reasonable selections then I'm concerned.  For clarity, I don't want Helix/LT due to size, weight & cost nor Hx Stomp that only has 3 footswitches - The Pod Go concept of an all in one gig worthy unit is perfect for me and I can manage with 4 assignable blocks - but I couldn't regularly manage 3 or less.  

 

This is a very important theme for me and hence I need to understand Pod Go's DSP limitations more clearly.  I've therefore changed the thread title slightly as I'd very much like to hear from users with Pod Go that have had an opportunity to spend time with it to see if/where they've encountered any DSP restriction issues in practice. 

 

 

 

 

Share this post


Link to post
Share on other sites
1 hour ago, voxman55 said:

.....I'm certainly not intending to use 3 reverbs. .....  Could I still be limited and cut down to e.g. only 3 assignable blocks? .....

 


Yes. It all depends on how DSP-intensive your selected 3 blocks are. As already described, 3 Reverb blocks probably precludes adding a 4th. Similarly, 3 other DSP-intensive blocks of any type may do the same.

 

1 hour ago, voxman55 said:

.....

If so I'd like to more clearly understand what combinations or models are particularly DSP hungry and which combinations could cause problems. If this is likely to be a very rare occurrence by choosing three of something that's one thing, but if it's going to trigger more regular issues on reasonable selections then I'm concerned. .....

 


I know there is a listing for Helix, submitted by a user, that identifies the relative DSP intensity of each amp, cab, and FX block. I don’t have an exact reference for you. I don’t believe such a listing has been created for the POD Go but it would be similar. In my experience with the POD Go I have never encountered the DSP limitation  (I.e. I’ve not seen any greyed out boxes) unless I am trying to double or triple up on some model types. I do not think this is a regular event with reasonable selections. You can count on being able to select all 4 of the assignable FX blocks in most cases.

Share this post


Link to post
Share on other sites

That's a little more reassuring @silverhead , thank you.

 

However I'd still very much like to hear from other Pod Go users to help me put this into a clearer perspective.  Whilst I appreciate manufacturers will always want to emphasise what their gear can do, I think it's also important to help customers better understand the potential ramifications of fundamental limitations of this nature. The manual comment referring to 3 reverbs is only there to evidence that Line 6 has raised the theme and is 'covered'.  But it's not really giving any meaningful information to its customers so that they have a frame of reference and can understand what percentage of DSP is used by each amp, cab, effect etc and thus where they might run into DSP limit problems.   

 

In the manual there are lists of models & FX .  Surely it should be possible to show what percentage of DSP would be used by each of these and to flag those that demand particularly high DSP percentage usage?  It really oughtn't to be up to users to have to try & figure this out themselves.   

Share this post


Link to post
Share on other sites

Well, I've been talking to folk about Line 6 DSP processing power on the Gear Page and they pointed me to this very informative breakdown of DSP usage for Line 6 Helix and Pod Go:

 

https://benvesco.com/store/helix-dsp-allocations/

 

Based on this, it looks like there could be some very worrying DSP restrictions. A Jumped Marshall SP100 model (I like Marshalls for classic rock stuff) uses over 40% of DSP over and above the 'semi-fixed block amp model 'allowance' - and even the non-jumped options for the same amp are around 35% (and that's without a cab!).  Most reverbs are around 34% and a good tape delay like the  Echoplex EP3 uses over 24% of DSP.   So you could struggle to even use just 2 of the flexible assigns!?

 

What I hadn't realised is that although there are 'semi fixed' and fully assignable blocks. that depending on the amp you use in the 'fixed' section, these fixed blocks are not 'ring-fenced' but directly impact on DSP availability elsewhere ie in the assignable blocks. So you could pick an amp, a delay and a reverb and have virtually nothing left - so instead of 4 assignable blocks you've been able to use only 2 before you run out. 

 

This means that you could regularly find restrictions very, very quickly which is worrying to say the least.  But in no video or review that I've seen has this been mentioned/discussed which I find odd.  So, is the chart accurate and is my understanding of how Pod Go allocates DSP processing correct?  

Share this post


Link to post
Share on other sites

This topic intrigues me so I spent my Sunday morning coffee time doing some exploration using my POD Go.

 

I set out to bring the POD Go to its knees in terms of DSP utilization using as few blocks as possible. I designed a preset that begins with the most DSP-intensive amp and Reverb models based on the above observations by voxman55. Having done that I then modified the preset by reducing DSP utilization in some model selections in order to permit the addition of certain model types to the preset until all 4 flex blocks are assignable. The details of the experiment are described in the attached document for interested readers.

 

For those who may only be interested in a summary....

The 'minimal' POD Go preset seems to be an amp with its default cab, the fixed FX blocks, plus 2 flex FX blocks (1 reverb and 1 delay). At this point, if you've selected the most DSP-intensive models of each of these types, you are unable to add a 3rd flex FX block to the preset. However, with some compromises you can readily edit the preset to achieve the allocation of all 4 flex FX blocks.

 

Also attached is my resulting preset. Note that I never even listened to this preset - I never had my guitar connected - so it likely sounds like cr@p. This was purely a preset design exercise using Pod Go Edit and the default parameters of all blocks. Listen at your own risk!!

 

EDIT: I just listened to this preset. I have 2 pieces of advice to you.

1) DON'T listen to this preset!!! If you choose to do so, turn DOWN your volume!

2) DON'T design presets by trying to max out the DSP!

 

And remember... this is an experiment designed to max out the POD Go's DSP. Under normal circumstances, without such nefarious intent, a user who did their research before purchasing the POD Go will rarely (imho) encounter meaningful DSP limitations.

 

Having said that, you might just be the user whose basic 'go to' preset is  the Brit Plexi Jump amp with the'63 Spring reverb and  6-head muli-tap delay. If that's the combination you can't live without - the POD Go is not for you unless you offload one of your reverb or delay FX to an external pedal in the FX Loop. Otherwise you should be OK.

 

POD Go DSP.txt

DSP Stress.pgp

Share this post


Link to post
Share on other sites

Hi Silverhead - many thanks for investing your time looking into this.  So, your investigation does seem to confirm that the Pod Go can indeed have unexpected DSP issues, and this can apply even with less DSP intensive amp models.

 

I'm into classic rock & blues and therefore use a number of Marshall tones. 


Amp
For me a Brit Plexi - Marshall Super Lead is pretty much a 'mainstay' amp for classic rock, I can select several options from a 'normal channel' variant that will use up 34.88% of DSP power, or I could go up to a Jumped version that will utilise 40.83% of DSP power. 

Cab
Cabs use no DSP over & above what's already allowed for unless (if I'm understanding the chart correctly) I want to import an IR, when it will use an extra 3.2% of DSP. 

So, I'll be reasonable here and forego the jumped model and go with the normal channel (If I want the bright channel it requires 35.73% of DSP) and I'll go with a stock cab with no imported IR.  Running DSP total: 34.88% 

Volume, Wah and EQ
Treated as requiring 0% DSP over the 'semi-fixed' block inbuilt DSP allowance and which I'm told utilise very little DSP in any event (so switching one or more of these off really won't release any meaningful DSP)  

Reverb
Next, I'd like to add a good old fashioned spring reverb. Both of the available spring reverbs utilise 34.03% of DSP.  I could go for a room, plate, chamber, hall etc at only 13.61% of DSP  but spring reverbs are pretty classic so let's say I'd like to stick with that.  Running DSP total: 68.91%

Delay
So, a nice tape-echo delay would be something like a Mastro Echoplex EP-3 which uses 24.24% of DSP
Running DSP total:  93.15%

Modulation

Lets go classic MXR phase 90, 5.10% of DSP. Running total: 98.25%

So far, so good.

Distortion boost
Let's stay pretty traditional and go for a mainstay, an Ibanez TS808 Tube screamer. 14.04%
Running DSP total: 112.29% - but hold on, that exceeds 100% DSP so I can't have it.

So I have a Marshall amp and cab, a spring reverb, a tape delay and an MXR Phase 90. Nothing exceptional.  The reverb, delay & phaser utilise 3 of the 4 fully assignable blocks - but I have nothing left for a tubescreamer unless I change the above selections for lower DSP models.  But if I go for higher DSP versions, then I could easily find myself having nothing left after 2 assignable blocks (as your test confirms).   

Going back to my TLSE/LE, and fully acknowledging that it's older modeling technology, I could still have all of this (Plexi, cab, spring reverb, tape echo, phaser & tubescreamer) with my TLSE/LE with no processing limit issues.  I mention this just to demonstrate that if you are considering changing your gear, 'dynamic DSP' means you may have to learn to think in a different way and you may be forced into making some unexpected compromises, as your own findings demonstrate.   And it's not just older gear like my TLLE/TLSE that doesn't use dynamic DSP, but new gear may not adopt this approach either e.g. Boss ME50/70/80 units.   

 

Observation

We are being asked by manufacturers to make a decision on spending not inconsiderable amounts of our hard earned money to buy their products. They know everything about their product, we know only what they tell us.  We might be able to pick-up further info and user tricks etc from reviews and user feedback & when it comes to sonic comparisons with competitors this is highly subjective and imported IR's and EQ can make huge differences anyway. So none of that 'subjective' stuff is an issue and you can't easily make it qualitative anyway. And feature comparisons are pretty much listed, reasonably easy to compare, and users can look at the spec and decide if they need 'tone capture' or whether 'snapshots' is better for their needs, or if they need an XLR out etc. I'm pretty relaxed here. 

 

But I am genuinely concerned where fundamentally important factual information such as DSP usage which impacts directly on customer expectation and 'the customer experience' is 'glossed over' by a single paragraph on p10 of the manual that simply says:

 

"Important: If you encounter items in the list that are greyed out or unavailable this means there isn't enough DSP to accommodate that category or model. For example, if you've selected three reverbs you probably won't be able to add a fourth". 

 

  • The DSP chart is not provided by Line 6 - this was only found following independent research. Without explaining which model uses which DSP how would users understand before buying that they might have some problems e.g. with their favourite selections?  
  • The above 'disclosure' is one small paragraph and refers to an unlikely selection choice which in my view gives the impression that users are unlikely to have a DSP issue unless they go for a pretty off the wall selection. Would buyers really have understood that selection of an amp model in the 'semi-fixed' block actually heavily reduces what's left to use elsewhere?  I certainly didn't - my initial thought was that amp/cab DSP was 'compartmentalised' and had nothing to do with the 4 flex blocks - re these. I thought  'that's OK, because I'm never going to select three reverbs!' 

 

When buying gear most of us will want more options, more flexibility and better tone - and manufacturers focus on this, which is why they list all the things that their latest 'mfx box' has in it and what it can do.  But it may not be immediately apparent to prospective buyers, and especially for those who are moving from older gear, that they might actually be restricted when they were not expecting to be. 

 

Please don't misunderstand me. I really like Line 6, I think they make great, innovative products, and I am in no way criticizing them specifically - I'm simply using them as an example only because I've specifically been looking at Pod Go and was very impressed with what I'd seen and heard - but was concerned following my research. Nor am I suggesting in any way shape or form that Pod Go isn't a great bit of competitively priced gear that isn't capable of producing some terrific tones that overall represents a very good gigging and home mfx solution. 

 

What I am suggesting is that with the growing complexity of products there should be greater responsibility on manufacturers to be a little bit more informative with regards to the product information they provide to us as consumers so that we are in a better position to make an informed decision.  Prospective customers should be able to more easily understand their products and be in a better position to compare with competitor products. DSP processing limitations is in my view an absolutely crucial theme.  

I'm not suggesting the industry needs to be 'regulated' (God forbid!!) but I do feel that customers and the industry would benefit greatly from some agreement led by the 'big boys' like Line 6 on 'best practice/fair customer charter' and that it should consider a way forward as an industry to convey this type of important information to its potential customers in an agreed format like the chart found.  It could avoid a lot of disappointment and gear returns too, which has a cost to retailers and manufacturers.  It shouldn't have to be up to enthusiasts to put this type of information together and for potential buyers to find this accidentally. It should be for manufacturers to provide this information so that consumers know where they are from outset.


 

Share this post


Link to post
Share on other sites

I understand that you are trying to make a product purchase decision and I am impressed with the level of detail and understanding that you have achieved in that effort. I am not trying to sway your decision one way or the other - just trying to help you get the facts and data behind some speculation and calculations.

 

I spent another while on this just to see what I could produce and to assess the accuracy of the DSP allocation chart produced by Ben Vasco based on your listed amp/FX blocks. It is very accurate. I was able to create the following preset: Brit Plexi Normal channel; 4x12 Greenback 25 cab; '63 Spring Reverb; Tape Echo; Script Phaser. According to your calculations and Ben's table that utilizes 98.25% of DSP. That seems to be very accurate because there's only 1 FX block I can now add as the 4th flex block from among all the FX types - the Hard Gate in the Dynamics section. According to the chart it utilizes 2.55% DSP which would exceed the 100% limit by 0.8%. The fact that it and only it can be added indicates that Ben's figures are very accurate. You can be confident using them to do further thought experiments on possible preset creations and limitations. If there's anything specific that you'd like me to test and verify just send me the preset specs and I will attempt to create it.

 

In particular, you are correct that the Tube Screamer cannot be added as the 4th block to hte above preset specification. You would have to make compromises to get it into your POD Go preset. For example, the Spring reverb is a very expensive choice in terms of DSP. Replacing that with either the Ganymede or the Glitz model allows you to add the Tube Screamer in the 4th block. Both the Ganymede and Glitz are very rich models and I am quite confident that you could choose one and set the parameters to achieve a result that you would like every bit as much as the Spring reverb. (Note: I tested but neither of these reverb models + Screamer allows you to restore the Brit Plexi Jump amp model that you would prefer.) The preset with 4 flex blocks is attached.

 

It seems the POD Go comes very close to being able to handle your ideal preset. You are making a small compromise in going with the Normal rather than Jump amp channel, and I expect you could find a satisfying compromise elsewhere to be able to use the Jump channel. I don't think you'll feel like you made a compromise with the reverb at all because  I expect you will be quite satisfied with the reverb model replacement; you may even prefer it.

 

A very positive note is that you are virtually fully utilizing the device DSP - which is a big part of the price/value equation. All devices that use a static rather than dynamic DSP allocation virtually guarantee that a significant amount of DSP remains unused and wasted in any preset. Wasted DSP is wasted $$. That's why I prefer dynamic DSP management; leave the user in control. Not to say that every preset you create should fully utilize the DSP but you are wise in making sure your purchase will be able to produce the specific amp and FX models that you want most to use.

 

Another thought.... If you really want to expand the DSP capacity of the device try using an external pedal in the FX Loop of the POD Go. In your case, purchasing a used Spring Reverb pedal would increase your available DSP by 25%. I know it's additional expense and I'm not recommending it, just noting the option.

 

Anyway, good luck with your decision. (The POD Go sounds excellent - just sayin')  ;-)

DSP Stress2.pgp

Share this post


Link to post
Share on other sites

Many thanks @silverhead, and I really appreciate all your input.  Thanks to the BenVisco info and your input, as well as other input gained elsewhere, I think I now have a much better understanding of how Pod Go works and am much more comfortable that it will meet my needs albeit I may need to do some 'creative tinkering' as needed. 

 

But with all respect to Line 6, I do feel they should enhance the manual to add a DSP usage figure for each model, with an explanation of Dynamic DSP and to explain that although there are 'semi-fixed' blocks these are not 'compartmentalised' i.e. the amp choices made in these, together with using an imported IR, can impact on the available DSP in the 4 flexible blocks.  I didn't understand all of this and I'm pretty sure there are a lot of other potential customers that are in a similar position.  Changing the example in their warning eg relating this to an actual amp, reverb, delay, fx combination that use higher DSP would also be very helpful and aid understanding . 

 

Share this post


Link to post
Share on other sites

Hi there! I made that DSP chart (that's my website). The chart is very accurate. The understandings discussed here are also generally correct. In practice, it's very subjective, but my advice is just use the dang thing without thinking about DSP at all. Even though I'm the one that literally made that chart, I never open it unless I hit a wall and run out of DSP (which happens rarely). In those cases I will open the chart and scope out what I can start to change. Maybe a "cheaper" delay or reverb will work for me without a noticeable change in sound.

 

The next piece of advice I'll give to anyone considering a purchase is not to focus on a specific real world amp model. Knowledge of amps is great and very helpful when tweaking. @voxman55, you've mentioned that the Brit Plexi is pretty much your dream amp but don't focus on that one amp model. In the world of amp modeling you may find that you get a much more pleasing "plexi" sound out of some completely different amp model. It may be surprising but it isn't bad or wrong. Don't spend time agonizing over how to get that one amp model just right. Try other amp models, too. That same advice works for other fx blocks as well. Spring reverb is awesome but do give the plate (and others) a try.

 

One correction is to make sure you are filtering and using the allocations chart to the fullest potential. There's a column "Type" which shows things like hx, L-s, or L-m where hx is a new effect from the HX line of modeling and L-s/L-m are "legacy" effects from the HD era (-s for stereo and -m for mono). That said, you wanted an EP-3 delay and there are two different EP-3 models. You did your calculations with the more expensive of the two. Same for the tube screamer which has two different models and you used the most expensive one. There's a search box above the chart which is interactive. Type in "screamer" and you'll see the two tube screamers side-by-side so you can more easily compare DSP usage.

 

 

  • Upvote 2

Share this post


Link to post
Share on other sites
9 hours ago, malhavok said:

Hi there! I made that DSP chart (that's my website). The chart is very accurate.......

 


Thank you for the chart and all the effort that went into it. During my experiments related to this thread I was amazed at the accuracy. The mere 0.8% discrepancy I noted at one point is essentially a rounding error and not indicative of any inaccuracy in your numbers. 
 

I strongly agree with you about how to make best use of the chart. Build your presets without any consideration for DSP until and unless you hit the limit. Then use the chart to determine the most DSP-effective adjustments to the preset.

  • Thanks 1

Share this post


Link to post
Share on other sites

Thank you guys, some good points there.  @malhavok - very impressive piece of work and tremendously helpful in helping my understanding.  

 

You'll be pleased to know that I'm not fixated on a single amp and I'm eager to experiment with Pod Go and discover all the things it can do. I picked the Plexi only because it was used by so many artists in the classic rock genre. 

 

As I said, I now have a better understanding of dynamic DSP and am now confident Pod Go will suit my needs.

 

Unlike a lot of younger folk (I'm 63 in August...Pod Go is my early birthday present to myself, assuming we get deliveries in the UK by then) , I was brought up on stomp boxes and then mfx units that didn't use dynamic DSP e.g. Yamaha GW33, Boss GT3, 5 anx 6, and Vox Tonelabs. My Zoom G5 is virtually immune from a DSP issue too. My Vox Valvetronix Amp and previous Line 6 Flextone II Plus were also static DSP gear.  So for me and others like me who haven't had previous experience of dynamic DSP it does mean a learning curve and a change in mindset.   In fact, until I started looking into all this, I'd never heard the terms static DSP or dynamic DSP.

 

I'm clearly getting old! 

Share this post


Link to post
Share on other sites
59 minutes ago, voxman55 said:

..... My Zoom G5 is virtually immune from a DSP issue too. My Vox Valvetronix Amp and previous Line 6 Flextone II Plus were also static DSP gear.  So if you haven't had previous experience of dynamic dsp it does mean a learning curve and a change in mindset. 

Yes, dynamic DSP management does expose the DSP limitations of a device. But that doesn’t mean that static management devices don’t have DSP issues, or are ‘immune’. Far from it. EVERY audio processing device has its DSP limitation.
 

All that static management does is ensure that you never encounter the limitation by placing handcuffs and constraints on the user. Your previous ZOOM, Valvetronix, and Flextone devices placed certain limitations on preset construction. E.g. only one reverb; no distortion and compression block in the same preset - choose one or the other, no stereo FX, no simultaneous mod and delay blocks, can’t have the Spring reverb at all - it’s simply not part of the product, etc. That way you were prevented from ever seeing the limitation. It also ensured that you could virtually NEVER use all the available DSP; there was always some waste. Of course the POD Go has its own limitations but not as constraining.

 

Dynamic management leaves you in control and provides much more flexibility while also permitting you to use virtually ALL available DSP when you want to. But yes, it takes user awareness and responsibility. Until you understand how it works it can leave the impression that the device is underpowered even though it is more powerful than its static management $$ competitors.
 

Enjoy your POD Go. I know you will.

Share this post


Link to post
Share on other sites

"Dynamic management leaves you in control and provides much more flexibility while also permitting you to use virtually ALL available DSP when you want to. But yes, it takes user awareness and responsibility. Until you understand how it works it can leave the impression that the device is underpowered even though it is more powerful than its static management $$ competitors."

 

Yup, I think that sums it up nicely! 

Share this post


Link to post
Share on other sites

I've just received my POD Go and like everything else, there are some compromises, but you just have to weigh up what's most important to you.  To me the POD Go is less of a compromise than the switching and block restriction in the Helix Stomp, and I just didn't want anything as big as the full blown Helix.  This lets me have a small board, and if I run out of DSP I've got a Tubescreamer and a fuzz sat in front of it and an M5 in the loop which pretty much means I can have any combination I'm likely to want, and the footprint and weight is still considerably less than the full Helix.  I'm really happy with it.

Share this post


Link to post
Share on other sites

I've had a couple of days now to start getting into my Pod Go. So far I've only been greyed out once and it was an easy tweak to compensate and still keep 4 blocks.  Just have to see how it goes as I start creating my own patches, and experiment. 

 

Have to say that I've been very impressed with the quality of all the amp, cab and fx models that I've tried so far.  This was a big factor when buying the Pod Go, so great job here Line 6. 

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...