Reviewer #1
Questions
2. I am an expert on the topic of the paper.
Strongly agree
3. The title and abstract reflect the content of the paper.
Strongly agree
4. The paper discusses, cites and compares with all relevant related work
Disagree
5. Please justify the previous choice (Required if “Strongly Disagree” or “Disagree” is chosen, otherwise write "n/a")
The paper cites relevant prior work, but it does not compare against the most relevant learned DJ-transition generation baseline, DJTransGAN. It compares only against Mosaikbox, a rule-based system. If DJTransGAN is not comparable due to task or cue-point assumptions, the paper should explicitly explain why.
6. Readability and paper organization: The writing and language are clear and structured in a logical manner.
Strongly agree
7. The paper adheres to ISMIR 2026 submission guidelines (uses the ISMIR 2026 template, has at most 6 pages of technical content followed by “n” pages of references, AI usage declaration or ethical considerations, references are well formatted). If you selected “No”, please explain the issue in your comments.
Yes
8. Relevance of the topic to ISMIR: The topic of the paper is relevant to the ISMIR community. Note that submissions of novel music-related topics, tasks, and applications are highly encouraged. If you think that the paper has merit but does not exactly match the topics of ISMIR, please do not simply reject the paper but instead communicate this to the Program Committee Chairs. Please do not penalize the paper when the proposed method can also be applied to non-music domains if it is shown to be useful in music domains.
Strongly agree
9. Scholarly/scientific quality: The content is scientifically correct.
Agree
10. Please justify the previous choice (Required if "Strongly Disagree" or "Disagree" is chosen, otherwise write "n/a")
ㄇㄥ
11. Novelty of the paper: The paper provides novel methods, applications, findings or results. Please do not narrowly view "novelty" as only new methods or theories. Papers proposing novel musical applications of existing methods from other research fields are considered novel at ISMIR conferences.
Agree
12. The paper provides all the necessary details or material to reproduce the results described in the paper. Keep in mind that ISMIR respects the diversity of academic disciplines, backgrounds, and approaches. Although ISMIR has a tradition of publishing open datasets and open-source projects to enhance the scientific reproducibility, ISMIR accepts submissions using proprietary datasets and implementations that are not sharable. Please do not simply reject the paper when proprietary datasets or implementations are used.
Agree
13. Pioneering proposals: This paper proposes a novel topic, task or application. Since this is intended to encourage brave new ideas and challenges, papers rated "Strongly Agree" and "Agree" can be highlighted, but please do not penalize papers rated "Disagree" or "Strongly Disagree". Keep in mind that it is often difficult to provide baseline comparisons for novel topics, tasks, or applications. If you think that the novelty is high but the evaluation is weak, please do not simply reject the paper but carefully assess the value of the paper for the community.
Disagree (Standard topic, task, or application)
14. Reusable insights: The paper provides reusable insights (i.e. the capacity to gain an accurate and deep understanding). Such insights may go beyond the scope of the paper, domain or application, in order to build up consistent knowledge across the MIR community.
Agree
15. Please explain your assessment of reusable insights in the paper.
The paper provides reusable insights about formulating DJ-transition generation as structured search over stem-wise gain trajectories. It also offers a useful example of separating internal objective diagnostics from perceptual evaluation, and shows how interpretable scoring terms can be used to analyze different aspects of transition quality. These insights are useful beyond the specific system.
16. AI Usage Policy: The paper complies with the ISMIR 2026 AI Usage Policy for Authors. (Authors are welcome to use any tool they wish for preparing and writing the paper but they must ensure that all content is correct and original. Authors must declare any use of LLMs and AI tools if it's a part of core methodology (including literature review, related work, generated figures, tables or other illustrations). Using AI tools for editing purposes (e.g,. checking grammar and fixing typos) need not be declared. Please pay attention to any hallucinated text, figures, or references and flag them in the comments below)
Agree
17. Please justify the previous choice (Required if "Disagree" is chosen, otherwise write "n/a")
n/a
18. Write ONE line (in your own words) with the main take-home message from the paper.
MixWeaver shows that DJ transitions can be generated by searching over interpretable stem-wise gain trajectories rather than directly predicting mixer controls.
21. Potential to generate discourse: The paper will generate discourse at the ISMIR conference or have a large influence/impact on the future of the ISMIR community.
Disagree
22. Overall evaluation: Keep in mind that minor flaws can be corrected, and should not be a reason to reject a paper. Please familiarize yourself with the reviewer guidelines at https://ismir.net/reviewer-guidelines
Weak accept
23. Main review and comments for the authors. Please summarize strengths and weaknesses of the paper. It is essential that you justify the reason for the overall evaluation score in detail. Keep in mind that belittling or sarcastic comments are not appropriate.
MixWeaver formulates automatic DJ-transition generation as stem-wise discrete optimization over an intro/outro overlap region. It separates tracks into stems, searches over stem-specific gain trajectories and transition windows, scores candidates with interpretable penalties, and selects a transition with a genetic algorithm. The system is evaluated with objective ablations and a blind listening study against Mosaikbox.

## Strengths

The paper addresses a real and underexplored generation problem, and the formulation is clear, interpretable, and complete. The system includes analysis, stem separation, candidate generation, scoring, search, rendering, ablations, and a human listening study. The listening evaluation is also relatively strong: blind A/B comparison, 23 participants including experienced DJs, randomized order, confidence intervals, participant-level paired tests, Holm correction, effect sizes, and honest reporting of the non-significant loudness-continuity result. The authors also frame the internal scoring ablation as a diagnostic rather than as direct perceptual evidence.

## Weaknesses

My main concern is that the evidence supports a narrower claim than the paper's framing. The paper shows that MixWeaver beats one rule-based baseline in an intro/outro smooth-blending setting, but it does not establish the stronger claims that structured interpretable optimization is competitive with learned DJ-transition generation, that stem-awareness is necessary, or that the objective reliably tracks perception.

1. **(Major) The most relevant learned baseline is missing.** The paper discusses DJTransGAN [1], but compares only against Mosaikbox. If [1] is not comparable, the paper should explicitly explain why; otherwise a matched-window comparison seems possible using the authors' own `fixed_window` setting.
2. **(Major) The objective is tied to smooth intro/outro blending, not general DJ transitions.** The gain-smoothness and energy-stability penalties penalize hard cuts, drop swaps, and abrupt handovers. This is plausible for intro/outro blending but not for arbitrary cue-point or hard-cut DJ practices.
3. **(Major) The evaluation lacks calibration baselines and anchors.** The objective experiments are explicitly restricted to intro/outro transitions, and the listening outputs appear to use the same current pipeline. This is exactly where a standard crossfade should be strongest, yet the listening test includes no pure no-stem/no-EQ raw-audio crossfade and no human-made DJ transition as a high anchor. The supplementary/code shows that `basic_swap_only` is a stem-based dB-linear crossfade-like condition, and the full system improves over it by only 0.06 on the objective, but this condition is never perceptually tested. Several load-bearing definitions, especially `basic_swap_only` and the template family, are left to the supplementary, making the ablations harder to interpret from the main text alone. Also, the paper cites ITU-R BS.1534-3 [2], but does not include reference/anchor conditions; if only the rating dimensions were adapted, this should be stated clearly.
4. **The stem-aware claim is not isolated.** Mosaikbox is a hybrid EQ+stem baseline, and all internal ablations remain inside the stem framework. A same-framework EQ-band ablation would be needed to show that Demucs stems provide a perceptual advantage over simpler frequency-band control.
5. **Objective-subjective consistency is not established.** The paper does not report correlations between objective scores and human judgments. Moreover, total-energy stability is one of the most load-bearing objective terms, while loudness continuity is the only perceptual dimension where MixWeaver does not outperform Mosaikbox.

## Suggestions

Add DJTransGAN [1] or explain why it is not comparable; add a standard raw-audio crossfade to the listening study; define `basic_swap_only` in the main text; add a same-framework EQ-band ablation; and report objective-human correlations, especially for energy stability versus loudness continuity.

## Overall assessment

MixWeaver is a complete and honestly evaluated system, and the Mosaikbox listening result is a real positive result. However, the core evidence is too narrow for the headline claims. The paper omits the closest learned baseline, lacks the key crossfade floor and human-DJ anchor, does not isolate stem-awareness, and does not establish that the objective tracks perception. I therefore recommend **Borderline / Weak Accept**.
Reviewer #2
Questions
2. I am an expert on the topic of the paper.
Agree
3. The title and abstract reflect the content of the paper.
Strongly agree
4. The paper discusses, cites and compares with all relevant related work
Strongly agree
5. Please justify the previous choice (Required if “Strongly Disagree” or “Disagree” is chosen, otherwise write "n/a")
"n/a"
6. Readability and paper organization: The writing and language are clear and structured in a logical manner.
Strongly agree
7. The paper adheres to ISMIR 2026 submission guidelines (uses the ISMIR 2026 template, has at most 6 pages of technical content followed by “n” pages of references, AI usage declaration or ethical considerations, references are well formatted). If you selected “No”, please explain the issue in your comments.
Yes
8. Relevance of the topic to ISMIR: The topic of the paper is relevant to the ISMIR community. Note that submissions of novel music-related topics, tasks, and applications are highly encouraged. If you think that the paper has merit but does not exactly match the topics of ISMIR, please do not simply reject the paper but instead communicate this to the Program Committee Chairs. Please do not penalize the paper when the proposed method can also be applied to non-music domains if it is shown to be useful in music domains.
Strongly agree
9. Scholarly/scientific quality: The content is scientifically correct.
Strongly agree
10. Please justify the previous choice (Required if "Strongly Disagree" or "Disagree" is chosen, otherwise write "n/a")
"n/a"
11. Novelty of the paper: The paper provides novel methods, applications, findings or results. Please do not narrowly view "novelty" as only new methods or theories. Papers proposing novel musical applications of existing methods from other research fields are considered novel at ISMIR conferences.
Strongly agree
12. The paper provides all the necessary details or material to reproduce the results described in the paper. Keep in mind that ISMIR respects the diversity of academic disciplines, backgrounds, and approaches. Although ISMIR has a tradition of publishing open datasets and open-source projects to enhance the scientific reproducibility, ISMIR accepts submissions using proprietary datasets and implementations that are not sharable. Please do not simply reject the paper when proprietary datasets or implementations are used.
Strongly agree
13. Pioneering proposals: This paper proposes a novel topic, task or application. Since this is intended to encourage brave new ideas and challenges, papers rated "Strongly Agree" and "Agree" can be highlighted, but please do not penalize papers rated "Disagree" or "Strongly Disagree". Keep in mind that it is often difficult to provide baseline comparisons for novel topics, tasks, or applications. If you think that the novelty is high but the evaluation is weak, please do not simply reject the paper but carefully assess the value of the paper for the community.
Disagree (Standard topic, task, or application)
14. Reusable insights: The paper provides reusable insights (i.e. the capacity to gain an accurate and deep understanding). Such insights may go beyond the scope of the paper, domain or application, in order to build up consistent knowledge across the MIR community.
Agree
15. Please explain your assessment of reusable insights in the paper.
Reframing the audio production task by optimization problem is elegant
16. AI Usage Policy: The paper complies with the ISMIR 2026 AI Usage Policy for Authors. (Authors are welcome to use any tool they wish for preparing and writing the paper but they must ensure that all content is correct and original. Authors must declare any use of LLMs and AI tools if it's a part of core methodology (including literature review, related work, generated figures, tables or other illustrations). Using AI tools for editing purposes (e.g,. checking grammar and fixing typos) need not be declared. Please pay attention to any hallucinated text, figures, or references and flag them in the comments below)
Agree
17. Please justify the previous choice (Required if "Disagree" is chosen, otherwise write "n/a")
"n/a"
18. Write ONE line (in your own words) with the main take-home message from the paper.
Reformulate DJ transition as optimization problem
21. Potential to generate discourse: The paper will generate discourse at the ISMIR conference or have a large influence/impact on the future of the ISMIR community.
Agree
22. Overall evaluation: Keep in mind that minor flaws can be corrected, and should not be a reason to reject a paper. Please familiarize yourself with the reviewer guidelines at https://ismir.net/reviewer-guidelines
Weak accept
23. Main review and comments for the authors. Please summarize strengths and weaknesses of the paper. It is essential that you justify the reason for the overall evaluation score in detail. Keep in mind that belittling or sarcastic comments are not appropriate.

Please find below my review


========================= REVIEW BEGINS HERE ===================================

Overall, this paper is well-written and well-organized. All relevant works are cited and clear. The core idea of this paper is to reformulate the DJ transition generation as the structured optimization problem. Authors propose that this formulation can help modeling the interaction between musical structure, stem-level coordination.

During reading, I have one confused term. What’s the difference between DJ transition generation and DJ mixing? In my understanding, the mixing term in DJ is a little bit different than traditional DAW-based multi-track mixing. This is because the DJ mixing is more like ``content creation’’, which means the processing is not only about the audio effects but also some ``generation’’, such as scratching or fancy sound effects.
After reading preliminary, I think the term ``generation’’ used in 3.1 is inproper.
That’s say, I think defining the term clearly is important for readers to capture the concept.


About the methodology part, I think the methodology is reasonable. The idea is to first construct the optimization space, this optimization space can be understood as space of gain curve and window (length, starting point). Then authors define the metric to cover three criteria: stem-overlap control, gain-curve smoothness, and energy stability. The design is good, but I only have one minor concern about the stem-overlap control. In this paper, the equalization is not considered, so if it is possible to include this will be very interesting (though I think it is non-trivial to include this).

About the evaluation, I appreciate authors conduct a comprehensive review. However, I still have some minor concerns. The first one is about the choice of baseline. The lack of some other baselines kinda let me be curious what’s the advantage of this optimization formulation compared to Chen et al. works [1]? Second, the evaluation dataset is vague. Though authors mention some statistics about the data, but what dataset? Is that the internal dataset or public dataset?


Though some minor concerns, I suggest to give the paper weak accept. This is because the methodology itself and the paper writing is all good in general sense.


Minor issues:
1. et al. -> \emph{et al.} for all?


[1] Chen, Bo-Yu, et al. "Automatic DJ transitions with differentiable audio effects and generative adversarial networks." ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2022.


========================= REVIEW ENDS HERE ===================================
Reviewer #3
Questions
2. I am an expert on the topic of the paper.
Agree
3. The title and abstract reflect the content of the paper.
Strongly agree
4. The paper discusses, cites and compares with all relevant related work
Agree
5. Please justify the previous choice (Required if “Strongly Disagree” or “Disagree” is chosen, otherwise write "n/a")
n/a
6. Readability and paper organization: The writing and language are clear and structured in a logical manner.
Strongly agree
7. The paper adheres to ISMIR 2026 submission guidelines (uses the ISMIR 2026 template, has at most 6 pages of technical content followed by “n” pages of references, AI usage declaration or ethical considerations, references are well formatted). If you selected “No”, please explain the issue in your comments.
Yes
8. Relevance of the topic to ISMIR: The topic of the paper is relevant to the ISMIR community. Note that submissions of novel music-related topics, tasks, and applications are highly encouraged. If you think that the paper has merit but does not exactly match the topics of ISMIR, please do not simply reject the paper but instead communicate this to the Program Committee Chairs. Please do not penalize the paper when the proposed method can also be applied to non-music domains if it is shown to be useful in music domains.
Strongly agree
9. Scholarly/scientific quality: The content is scientifically correct.
Strongly agree
10. Please justify the previous choice (Required if "Strongly Disagree" or "Disagree" is chosen, otherwise write "n/a")
n/a
11. Novelty of the paper: The paper provides novel methods, applications, findings or results. Please do not narrowly view "novelty" as only new methods or theories. Papers proposing novel musical applications of existing methods from other research fields are considered novel at ISMIR conferences.
Agree
12. The paper provides all the necessary details or material to reproduce the results described in the paper. Keep in mind that ISMIR respects the diversity of academic disciplines, backgrounds, and approaches. Although ISMIR has a tradition of publishing open datasets and open-source projects to enhance the scientific reproducibility, ISMIR accepts submissions using proprietary datasets and implementations that are not sharable. Please do not simply reject the paper when proprietary datasets or implementations are used.
Agree
13. Pioneering proposals: This paper proposes a novel topic, task or application. Since this is intended to encourage brave new ideas and challenges, papers rated "Strongly Agree" and "Agree" can be highlighted, but please do not penalize papers rated "Disagree" or "Strongly Disagree". Keep in mind that it is often difficult to provide baseline comparisons for novel topics, tasks, or applications. If you think that the novelty is high but the evaluation is weak, please do not simply reject the paper but carefully assess the value of the paper for the community.
Agree (Novel topic, task, or application)
14. Reusable insights: The paper provides reusable insights (i.e. the capacity to gain an accurate and deep understanding). Such insights may go beyond the scope of the paper, domain or application, in order to build up consistent knowledge across the MIR community.
Agree
15. Please explain your assessment of reusable insights in the paper.
Reusable insights include the value of a stem-aware mixing approach and a scoring function composed of different evaluation metrics.
16. AI Usage Policy: The paper complies with the ISMIR 2026 AI Usage Policy for Authors. (Authors are welcome to use any tool they wish for preparing and writing the paper but they must ensure that all content is correct and original. Authors must declare any use of LLMs and AI tools if it's a part of core methodology (including literature review, related work, generated figures, tables or other illustrations). Using AI tools for editing purposes (e.g,. checking grammar and fixing typos) need not be declared. Please pay attention to any hallucinated text, figures, or references and flag them in the comments below)
Agree
17. Please justify the previous choice (Required if "Disagree" is chosen, otherwise write "n/a")
n/a
18. Write ONE line (in your own words) with the main take-home message from the paper.
The paper presents a novel approach to mixing tracks that (1) navigates the transition space using a scoring function composed of multiple metrics and (2) creates transitions for stems separately.
21. Potential to generate discourse: The paper will generate discourse at the ISMIR conference or have a large influence/impact on the future of the ISMIR community.
Disagree
22. Overall evaluation: Keep in mind that minor flaws can be corrected, and should not be a reason to reject a paper. Please familiarize yourself with the reviewer guidelines at https://ismir.net/reviewer-guidelines
Weak accept
23. Main review and comments for the authors. Please summarize strengths and weaknesses of the paper. It is essential that you justify the reason for the overall evaluation score in detail. Keep in mind that belittling or sarcastic comments are not appropriate.
The strengths for me are the establishment of a interpretable mixing system, the ablation studies showing the value of different scoring function components, the samples provided along with the paper, a clear problem solution, and human evaluation. However, the evaluation does not indicate how the method compares to previous methods, making it harder to assess the value of the paper.