Forum Replies Created

Viewing 15 posts - 91 through 105 (of 367 total)
    • 7 May 2025 at 7:30 am #1031

      Dual-column files are assumed to be impulse response time samples, where the 1st column is the timestamp and the 2nd samples, which is why they aren’t importing correctly for you. See

      Loudspeaker Measurement Import File Formats

      The quickest solution I can suggest is to load the TXT file Excel or Google Sheets, add a 3rd column of all zeros (for phase), then export as a new TXT file or CSV. (You might need to select just the active cells, prior to export.)

      If that doesn’t work, email your TXT file to info ‘at’ <our website> and we’ll take a look.

      Regards,
      EA Support

      I tried to attach one of the files FYI, but this portal doesn’t seem to allow it

      FWIW you can put the file on one of the file-sharing websites, or dropBox, and include a link in the forum message.

        30 April 2025 at 6:26 pm #1028

        ok, thank you for quick reply
        We use usually rephase, but for a project we need chebyshev filter type 1 for crossover use, in minimum phase FIR format,, it is possible?

        Yes, to create a Chebyshev equivalent FIR filter, use the ‘FIR: Magnitude Adjust’ tab in FIR Creator EX. You can try this in the demo mode. However, since you mentioned minimum-phase, it’s much more efficient to implement this using IIR biquads.

        For export IIR coefficient we need FIR Designer or Creator is ok?

        The FIR Creator EX ‘IIR Filters’ tab can be used to generate biquad coefficients, which can be exported from that tab.

        Please download the demo version and experiment with all the tabs in the workflow.

        To learn more about the workflow, see the first two ‘FIR Designer’ videos on:
        https://eclipseaudio.com/fir-filter-design-tutorial/

        About audio quality, are there differences in quality for the same filter between Fir Creator and Rephase?
        Thank again
        Best regards

        No. The perceived ‘quality’ of FIR and IIR filter processing relates to the design choices in creating the filters, not the choice of program.

        Regards,
        EA Support

          23 April 2025 at 12:28 pm #1026

          Yes the FIR filter needs to be designed and exported for the sample rate that the processor/amp uses for FIR filters. (I don’t know the specifics w.r.t. Linea Research amplifiers.)

          Some 96kHz processors run their FIR filters at 48 kHz, so the FIR filter needs to be designed for 48 kHz (and not the core 96 kHz). And Lake uses 12kHz, 24kHz and 48 kHz for FIR filters, even though the input processing runs at 96 kHz.

          In our products, if you have a design created for a particular sample rate, you can easily export for a different sample rate by first changing the “design sample rate”.

          Regards,
          EA Support

            20 April 2025 at 9:11 pm #1024

            We have many things we are working on so we can’t give an ETA at this time.

            Regards,
            EA Support

              29 March 2025 at 7:39 pm #1017

              Thanks very much!

                28 March 2025 at 6:08 am #1015

                >> My backup plan is to use the latest IIR presets, which I have used for years and
                >> like the sound, feed them into FIR Designer-M, and create new, similar, but phase-linear versions.

                You can recreate the presets from electrical measurements, however I would NOT convert each channel to linear-phase. The phase differences between channels in the multi-way are part of the preset and are inherent in making the preset work. If you change these relationships, you could dramatically alter the system’s response and directivity in and around the crossover region.

                As for making the combined acoustic response linear-phase, in addition to electrical measurements of the preset, you would need some reliable or meaningful acoustic measurements.

                Regards,
                EA Support

                  27 March 2025 at 7:21 am #1013

                  The SPS files appear to be proprietary binary preset files. We can’t load or open these.

                  You could try loading each of these presets into a EV processor, then turn off all the IIR filters and crossovers (leaving just the FIR filters running) and then electrically measure the FIR filters at the outputs of the processor. From this, you can use the process described above to generate new FIR filters.

                  We’re a small company and don’t have the resources for individual phone support. If you have a measurement, we can show you the workflow (with screenshots) to then do more conversions yourself, but it sounds like the main issue right now is actually getting at or measuring the FIR filters.

                  Regards,
                  EA Support

                    27 March 2025 at 7:14 am #1012

                    Hi Eddi,

                    ‘Phase only’ on import: We’ve added that to the todo list.

                    Two measurement paths – separate magnitude and phase: It’s unlikely we’ll implement this but you can achieve this in the following way.
                    – Create your measurement as a 3 column TXT or CSV file (freq Hz, mag dB, phase deg) with the mag and phase from your separate measurements
                    – Load this measurement on the Import tab
                    – Do all your desired magnitude manipulation on the Mag Adjust and Auto Mag tabs, and make sure ALL the filters are set to ‘Linear phase’
                    – Do all your desired phase adjustments on the Phase Adjust and Auto Phase tabs
                    The result at the Voicing tab is the combination of the above, and you can then voice to taste.

                    Keyboard keys for filter dB, Q and frequency: We’ll consider it.

                    Regarding the binaural measurements, are you request we add measurement capability into our software? We have considered this but many people are very particular about the measurement tools they use – Klippel, SMAART, SysTune, SoundCheck, REW, etc. – so to date, we haven’t ventured into this space.

                    Regarding the ‘AVG + phase from one ear’, it sounds like you would like an additional “Averaging Mode” called “Power (using Ref. phase)”. With this, the phase from the measurement denoted as the ‘Reference’ would be used. We’ll consider this.

                    Regards,
                    EA Support

                      24 March 2025 at 6:53 am #1006

                      We are considering releasing a direct conversion tool, however you can get an almost identical result with our tools by loading an EV FIR filter on the channel target tab. The workflow is described in this post:

                      Converting FIR filter from 48 kHz to 96 kHz

                      With this approach you can also create the new FIR filters at different sample rates. And, ensure the effective filter delay (ms) is the same as the old filters, so that you don’t introduce timing differences between old and new platforms.

                      >> Will FIR Designer-M accomplish this?

                      Yes, as above. Make sure the system target is turned off before working in the channel workflow. FIR Designer (single channel) can also do this, but with FIR Designer M you get the ability to design and simulate full multi-way designs.

                      Regards,
                      EA Support

                        6 March 2025 at 5:58 pm #1003

                        Noted.

                        Thanks,
                        EA Support

                          26 February 2025 at 8:22 am #1001

                          We have looked through the FIR Designer M multi-band limiter code in depth and can’t see any problems. However one idea comes to mind.

                          The three bands of the multi-band limiter are, by default, un-muted. If multi-band limiter crossovers have not yet been set, then all three bands will be passing audio and summing. This equates to 9.54 dB of gain.

                          Try muting all but one band. This should bring the gain back to 0 dB.

                          Regards,
                          EA Support

                            25 February 2025 at 9:21 pm #999

                            Hi LB,

                            We don’t currently have any Lake products set up here to test, so it will take some time. In the meantime, try setting the Lake Controller in Designer mode (in User Preferences), then for the XP module, go to “IO Config and Worksheets” “Levels” tab, and scroll vertically through the table. Various settings, including for the MBL, are displayed.

                            What version of the Lake Controller are you using?

                            Regards,
                            EA Support

                              2 February 2025 at 11:26 am #991

                              Yes, the filter length is the number of samples or taps. And yes, the delay allows you to move the peak within the filter length. The delay represents the time=0 point in the filter and is the effective bulk processing delay through the filter. The delay is not needed for SigmaStudio.

                              Please read the FIR filter guide to learn more.

                              FIR Filter Guide

                              Regards,
                              EA Support

                                18 January 2025 at 4:01 pm #988

                                You can check these yourself by loading them into the Import tab.

                                FIRLR2 sample1_1.bin & FIRLR2 sample1_3.bin are the HF (channel 3)
                                FIRLR2 sample1_2.bin is the MF (channel 2)

                                >> Probably there is some bug in your software.

                                We just checked the export for each channel here and it’s working correctly.

                                Regards,
                                EA Support

                                  16 January 2025 at 6:48 am #983

                                  Do any of the Averaging options take into consideration the Coherence of the Transfer Function measurements in use?

                                  No, the coherence is not used. However, incorporating coherence in the averaging process is on the TODO to investigate and implement.

                                  Regards,
                                  EA Support

                                Viewing 15 posts - 91 through 105 (of 367 total)