Yeah thats a tricky one. Personally, I never have used the vertical zoom on the analog channels. Most of the time I only display the digital ones (especially if I have all 16 channels active). Why not use Shift + mouse wheel to vertical zoom in and out and Ctrl + mouse wheel to scroll up and down? You wouldn’t have to change your hand position because both keys are on top of each other.
1 Like
I have a suggestion in regards of the timing markers and their text which you can add.
If you write a long comment to a marker (for example: Start of internal read cycle. IC retrieves a data byte from an internal register pointed by the adress lines [A2:A0]) it isn’t shown fully in the Timing Markers tab. To read the text you always have to navigate to the three small dots and click on Edit or hover above the marker where the signals are being displayed.
Now my suggestion: Why not use the empty space under it to show the full comment?
See attached screenshot for a rough idea.
1 Like
Another suggestion to the timing markers:
The vertical lines of the timing markers are hard to make out because they are so transparent especially if they cross a lot of signals.
Would it be possible to make them a little bit stronger?
See attached screenshot
1 Like
Oh answer was literally under my nose. I almost never encountered the 8-bit, read/write bit included interpretation method thanks for the explanation.
I’m used to read addresses as 8-bite, rw bit set as 0 but I checked documentation of our actual peripherals and none of the two interpretation is more used than the other.
By the way, great work on this software redesign. Do you plan to use the frame in your Analyzer SDK? I’ve worked on a complete ISO-7816 Analyzer i want to share but i was waiting for frame implementation for better decoding 
1 Like
That’s pretty much the plan
1 Like
When you hover over the note it’ll expand the text, and clicking should open the edit.
We’ll add that soon.
Hello! I just started using a Saleae Logic 8 this week; very happy so far. I am also starting to write my own analyzers for some projects.
I notice in Logic 1.x if I assign an analyzer to a channel and then start simulation, the analyzer’s simulated data is put in the channel and decoded.
However in the Logic 2.x betas this doesn’t appear to happen. I always seem to get the same kind of random simulated square wave data no matter what analyzer I choose. This seems like a bug?
See for example this simulated run with an I2C analyzer on Channel 0 and 1. Definitely not I2C-like signals.
First of all, happy to hear that you’re happy so far
You’re right, Logic 2 doesn’t support protocols simulation at the moment, and it always produces the same “random” data. Is it a deal breaker for you? How do you use the simulated data usually?
It’s a showstopper for developing analyzers, certainly.
But otherwise, I really enjoy seeing the simulated data from an educational perspective to see how well formed signals look for different protocols. Plus, it was a huge selling point before I bought the actual hardware that I could simulate protocols and see how the software actually works. If I couldn’t have done that I’m not sure I would have been convinced to buy a Logic.
1 Like
These are valid points
We’re planning on improving (and simplifying) the analyzers development process, and I believe that we’ll support simulation as part of that. In the meantime, I added your request to our new feature requests board:
https://ideas2.saleae.com/b/feature-requests/simulating-valid-protocol-decoders-data
By the way, we’ll have high-level Python analyzers in a couple of weeks. I’d love to get your feedback on that as someone who built analyzers in the past.
Just wanted to follow up and mention that I’m using 2.2.1 and 2.2.2 successfully now on multiple monitors. The change i made was switching the graphics mode in the bios so it’s now using the dedicated gpu and nvidia drivers. Prior to this it was using the combo intel graphics on my lenovo X1 extreme.
1 Like
Hello, I used the Alpha a few days ago and I have some feedback:
- When you have some decoded data, when zooming out the text tends to disappear even if there’s enough space to display it. It looks like there’s some margin set, so when text+margin doesn’t fit the blue rectangle the text disappears.
- I couldn’t find a way to search for a byte pattern. I’m confused because there’s an old proposal here:
https://ideas.saleae.com/ideas/SALEAE-I-406#1
Is this still not supported?
I’m happy to hear that!
Thanks for letting us know (we’ll try to reproduce it here)
First of all, thanks for the feedback!
text tends to disappear
Does it disappear completely? The app should display the version that fits best to the size of the box.
If you can share a screenshot that would be great!
I couldn’t find a way to search for a byte pattern.
We don’t support it yet but we’re already working on an alternative:
- You’ll be able to run a simple high-level analyzer that combines results (into packets for example).
- Search for the combined value
What do you think?
Analyzer data visualization
I’m not at my work PC, so I can’t double-check there unfortunately.
But, in general, the representation of data can be improved. For instance, here’s a quick screenshot from my Mac (analyzer processing random data from demo mode):
My observations here:
- There is no margin, so the text can extend to the edge of the blue rectangle, which is good. Is it the same between Win and Mac? I thought it would not display anything in a box if the full text doesn’t fit.
- The “0x” prefix burns 2/3 of the space. I would just remove it when the mode is Hexadecimal, there’s no need to distinguish from other bases. Or, if you think it could be confusing when you have other traces on different modes, you could write it on top:
So, in general, it’s a matter of using the space efficiently.
As a side note, it looks like on macOS when I zoom using the scroll wheel, it zooms 3 steps at the time. It goes from very zoomed in to very zoomed out. I’m pretty sure it wasn’t doing this on Windows.
Data search
Regarding the search, it would be nice to stack different analyzers and be able to search at whatever layer you want. For instance, an I2C analyzer could have an Eeprom analyzer on top which presents the read/write operations.
I understand this feature is not trivial, because on some protocols you may have special symbols, like Start-Of-Frame, which is not a byte. But usually we’re dealing with bytes, so it’s a good starting point.
Here’s an idea regarding the format for the search expression:
Searching for a few bytes in sequence: 0x55 0x12 0x34 0x56
Searching with a wildcard (any byte matches): 0x55 ? 0x34 0x56
Searching for a few bytes in binary: 0b01010101 0b00001111 0b11110000 0b00000000
Searching for a byte followed by an ASCII string: 0x55 "HELLO"
Searching with mixed formats: 0x55 0b01100110 "HELLO"
This would take care of most cases and would be a major step forward.
If you want to get more fancy, here are more ideas:
Searching with special symbols or bus states (not bytes): $SOF 0x11 0x22 0x33 $EOF
Searching with wildcards on byte value: 0x55 0x4? 0x11 0b??000000
Searching for multi-byte values (big endian): 0x12345678
Searching for multi-byte values (little endian): 0x12345678:LE
For weird stuff, it would be ok to have a checkbox near the search textbox to switch to a text search on the analyzer output text (which is what Logic is doing now AFAICT).
Ah, one final thing. It would be cool to highlight matches in the timeline. So, if I search for some syncronization pattern, I’d like the matching bytes to be displayed in boxes with a different color instead of blue (say, orange or whatever).
First of all, thanks for the detailed feedback! It’s super helpful 
There is no margin, so the text can extend to the edge of the blue rectangle, which is good. Is it the same between Win and Mac?
Yes, we decided to remove any margins to maximize the visible text.
The “0x” prefix burns 2/3 of the space.
You’re right. We’re planning on revisiting the rendering process of analyzer results and we’ll fix that as part of it.
As a side note, it looks like on macOS when I zoom using the scroll wheel, it zooms 3 steps at the time.
Could it be related to your mouse or your settings? It works fine on 3 Macs that we have here (sorry for not being more helpful here)
Data search
We’re currently working on high-level analyzers that we’ll allow you to add more layers on top of the existing analyzers. As part of this, you’d be able to combine results and by that, search for them together. Would it be helpful?
Implement multi-row search in databases is not trivial at all, and we believe that with high-level analyzers you’ll have more control over the output.
What do you think? Would that be a good solution?
Ah, one final thing. It would be cool to highlight matches in the timeline
We’ll look into this, that’s a cool idea. Our plan right now is to (optionally) scroll the table on the sidebar to match the graph when you’re panning left/right.
Thanks again