{"id":"cb0527cb-98bd-49d2-b85e-6ade922b6cae","arxiv_id":"2501.17872","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A simple doublet lens is ray-traced with KrakenOS, its PSFs are convolved onto Woodscape fisheye simulations, and the resulting edge blur is measured, showing stronger degradation toward the image periphery.","lead":"The authors build a pipeline that degrades synthetic fisheye driving images using optical blur from a two-glass lens simulated in the open-source ray tracer KrakenOS, and they compare that lens simulation against the commercial Zemax tool. The work is useful for making automotive simulation data look more like real camera output, though the lens used is a simple doublet, not an actual fisheye lens.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The realism claim is not supported: a ±2° doublet PSF grid is physically mismatched to a ~180° fisheye image, and the paper's own Limitations concede this; the conditional verdict should stand.","rationale":"The central claim is two-part: (1) KrakenOS matches Zemax on the doublet, and (2) convolving the resulting PSF grids onto fisheye simulations produces realistic spatially varying optical artifacts. The first part is supported by RMSE ~0.02 for the three RGB wavelengths at 0° and ±2°, though it is a narrow sample. The second part is where the argument breaks. To establish realism, the PSF at each image location must correspond to the optical system and incident angle that produced that location. Here the image is produced by a fisheye lens with large FOV and strong distortion, but the PSF model is a doublet sampled only over ±2°. Step 4 resizes the 8×8 grid to 8×6 or 32×24 and convolves each patch with a PSF indexed by patch location; there is no mapping from image pixel to the doublet's object-space angle that would make this a physically correct degradation. Consequently, the demonstration that MTF50 drops more in middle/edge regions than center is consistent with a spatially varying convolution, but not uniquely with realistic fisheye-lens aberrations; it could be an artifact of the chosen PSF grid placement. The authors' Limitations sentence is a direct, manuscript-internal admission of this gap. The paper remains a reasonable proof-of-concept for an open-source ray-tracing degradation pipeline and a useful cross-tool PSF comparison, but the strong realism claim in the abstract is not established. A wide-angle lens model covering the actual FOV and an evaluation on controlled slanted-edge targets would settle the concern. No independent code/data release is provided, so the RMSE comparison itself is not independently reproducible; however, the primary load-bearing issue is the physical mismatch, not reproducibility. The Pith Reader identified this mismatch as the weakest assumption and issued CONDITIONAL; I agree. The verdict should remain CONDITIONAL, since the tool comparison and proof-of-concept are still worth publishing with the realism claim appropriately qualified.","tokens_in":7072,"tokens_out":6179,"duration_ms":68589,"concrete_test":"Re-run the pipeline with a wide-angle fisheye lens model in KrakenOS spanning the full MVL FOV, generating PSFs at the actual object-space angles corresponding to each fisheye pixel, and convolve the same two source images; recompute the MTF50 radial profiles. If the center/periphery sharpness pattern is materially different from the doublet-based result, the original realism claim is not supported; if it closely matches, then the doublet approximation was adequate for the tested images.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is the transfer from the ±2° doublet PSF computation (Methodology step 3) to the fisheye image convolution (Methodology step 4). The PSF grid is sampled in object-space field angle for a doublet that is not the Woodscape fisheye lens, while the MVL fisheye images span a much larger angular range under a different, strongly nonlinear projection. Assigning each image patch a PSF purely by its pixel position therefore has no physical correspondence to the angles at which light actually reached the fisheye camera. The observed center-to-edge MTF50 decrease could be generated by the arbitrary patch-to-PSF mapping rather than by field-dependent aberrations of a real fisheye lens. The authors explicitly concede this: 'The degradation itself is not realistic because the warped simulation is convolved with a completely different lens' (Limitations section). Since the abstract and contributions claim 'realistic optical artifacts,' this concession removes the main realistic-degradation support; the remaining tool comparison (RMSE ~0.02 at only nine field/wavelength combinations) is a useful but narrower validation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes SOLAS, a pipeline that uses the open-source Python ray-tracing library KrakenOS to generate point-spread-function (PSF) grids for an optical doublet and convolves them with synthetic fisheye automotive images from SynWoodscape and Parallel Domain Woodscape, with the goal of introducing realistic optical aberrations into simulation. The authors compare KrakenOS PSFs against Ansys Zemax OpticStudio at three field angles and three wavelengths, reporting an average RMSE of about 0.02, and they measure image sharpness before and after degradation using the ISO 12233 slanted-edge method, reporting that edge and middle regions lose contrast while the center does not. The paper claims that KrakenOS can match industrial ray-tracing performance and that the proposed pipeline introduces spatially varying optical artifacts into fisheye simulations.","tokens_in":7279,"tokens_out":2264,"duration_ms":23603,"significance":"If the central claims were fully supported, the paper would make a useful contribution: an open-source alternative to commercial ray-tracing for automotive simulation, a concrete pipeline for PSF-based degradation of synthetic fisheye data, and a quantitative sharpness evaluation. The external benchmark against Zemax and the use of an independent MTF50 metric are strengths, as is the explicit code-adjacent methodology that enables replication. However, the paper's headline claim of 'realistic optical artifacts' is explicitly contradicted by its own Limitations section, and the restricted scope of the validation (three field angles for a ±2° doublet applied to ~180° fisheye images) means the practical significance for automotive perception is not established. As a proof-of-concept for degrading fisheye simulations with a generic doublet, the pipeline is demonstrated; as a claim of realistic fisheye degradation, it falls short.","major_comments":[{"comment":"The abstract states the pipeline introduces 'realistic optical artifacts' into fisheye simulations, and the contributions claim a 'novel optical degradation pipeline' with 'realistic' degradations, but the Limitations section explicitly concedes: 'The degradation itself is not realistic because the warped simulation is convolved with a completely different lens.' This is a load-bearing contradiction: the central realism claim is unsupported. The authors should either reframe the contribution as a proof-of-concept for a generic degradation pipeline, or replace the doublet with a fisheye-appropriate lens model and recompute the PSF grids for the actual field angles of the Woodscape MVL images.","section":"Abstract and Limitations"},{"comment":"The RMSE comparison between KrakenOS and Zemax is reported for only three field angles (0°, −2°, +2°) and three wavelengths, with no description of how the PSFs were normalized, resampled, or aligned before computing RMSE. The statement 'KrakenOS can match industrial performance' is therefore not established by the presented evidence, because a raw RMSE value of 0.02 depends entirely on the PSF normalization convention (e.g., peak value, total energy, or unit-integral scaling) and on the pixel grid alignment. The authors should specify the exact comparison methodology, report RMSE as a function of PSF normalization, and include a figure or table showing the per-pixel difference maps.","section":"Table 2 and Section 'Build & Finetune KrakenOS'"},{"comment":"The load-bearing transfer from the ±2° doublet PSF computation to the fisheye image convolution is physically mismatched: the PSF grid is sampled in object-space field angle for a doublet that is not the Woodscape fisheye lens, while the MVL fisheye images span a much larger angular range under a strongly nonlinear projection. Assigning each image patch a PSF purely by its pixel position therefore has no direct physical correspondence to the angles at which light reached the fisheye camera. The observed center-to-edge MTF50 decrease could be generated by the arbitrary patch-to-PSF mapping rather than by field-dependent aberrations of a real fisheye lens. The authors should either (a) simulate a fisheye lens whose distortion and field map match the Woodscape camera, or (b) clearly label the experiment as a test of the degradation pipeline with a deliberately mismatched lens and discuss how the results would change with a matched lens.","section":"Methodology steps 3–4 and Figure 5"}],"minor_comments":[{"comment":"The notation for surface labels in Table 1 (e.g., 'POb j' and 'PI ma') appears to contain word fragments; please replace these with standard abbreviations such as 'Object Plane' and 'Image Plane' for clarity.","section":"Methodology step 3"},{"comment":"The caption describes '8 × 8' and '32 × 32' grid sizes, but the text in Methodology step 4 states the images are 1280×960 and the grids are resized to 8×6 and 32×24; please harmonize the caption with the actual grid dimensions.","section":"Figure 2 caption"},{"comment":"The sentence 'An image was chosen of one camera position from both the SynWoodscape [10], and Parallel Domain Woodscape [11] datasets' is grammatically unclear; it should state that one MVL image was selected from each dataset.","section":"Related Work"},{"comment":"The qualitative claim that the degraded images show 'coma and astigmatism towards the periphery' is based on visual inspection of Figure 4, but no direct PSF shape analysis at the corresponding image coordinates is provided; adding a panel that overlays the local PSF at a periphery location would strengthen the qualitative argument.","section":"Section 'Optical Degradations of Fisheye Automotive Simulation'"},{"comment":"The sentence 'The optical doublet, a key lens component of optical systems (including fisheye), can be replaced for more complex optical systems in our pipeline' is ambiguous: it should be clarified whether 'replaced' means 'substituted with' a more complex lens or 'used in place of' one.","section":"Conclusion"}],"recommendation":"major_revision","confidential_remarks":"The paper is a short engineering report that would benefit from reframing its contribution as a proof-of-concept for PSF-based degradation rather than claiming realistic fisheye degradation. The self-concession in the Limitations section is the main blocker; if the authors revise the claims and add the missing comparison methodology details, the paper could be suitable for a workshop or a short conference paper. The scope of the journal may also be a consideration: the paper's focus is primarily on an open-source ray-tracing tool and an image-degradation pipeline, not on a new signal-processing or optics theory result."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the useful result is the RMSE comparison between KrakenOS and Zemax on a simple doublet (around 0.02), and the demonstration of a spatially varying PSF convolution pipeline for automotive fisheye simulation. The paper is honest about its main limitation, which is a good thing, but that limitation undercuts the headline claim of 'realistic optical artifacts.'\n\nWhat's new: applying a PSF grid generated by KrakenOS to SynWoodscape and Parallel Domain Woodscape fisheye images, then measuring MTF50 with ISO12233 slanted-edge method. That combination is new, and the pipeline is clearly described. The tool comparison is the strongest part: nine PSF pairs at three field angles and three wavelengths, RMSE around 0.02, with blue slightly worse. That's a reasonable initial validation of an open-source tool against an industry standard.\n\nSoft spots: the load-bearing issue is the FOV mismatch. The doublet is simulated at +/-2 degrees, the fisheye images cover roughly 180 degrees, and the PSF grid is tiled by pixel position. There is no physical mapping from the doublet's object-space angles to the fisheye's projection. The authors concede this in Limitations: 'The degradation itself is not realistic because the warped simulation is convolved with a completely different lens.' That concession directly contradicts the abstract's 'realistic optical artifacts.' So the central realism claim is not supported. What remains is a proof-of-concept for spatially variant degradation, which is fine if framed that way.\n\nAlso: MTF50 numbers come from a single image per dataset, no error bars, and the PSF normalization and convolution details are underspecified. The 32x24 grid uses 40x40 regions per PSF, which is a very coarse approximation of a PSF. That's fine for demonstration but not for claiming high fidelity. And there's no code or data release, which would help reproducibility.\n\nOn the citation pattern: prior work (Heide et al., Yang et al., Wittpahl et al.) is properly cited for PSF-based degradation. The self-citations to SS-SFR are relevant; not a problem.\n\nBottom line: this is a workshop-level proof-of-concept, not a fully supported realism result. The tool validation is useful and the pipeline is a clear starting point. I'd send it to peer review with a request for major revision: match the lens model to the camera FOV, add multi-image evaluation with error bars, release code/data, and soften the realism claim. A serious referee would get value from engaging with it.","headline":"A useful validation of KrakenOS against Zemax, but the realism claim for the fisheye degradation is explicitly conceded by the authors; treat this as a proof-of-concept pipeline paper, not a realism result.","tokens_in":7863,"tokens_out":2314,"would_cite":false,"duration_ms":23578,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that an open-source ray tracer, KrakenOS, produces point spread functions matching Ansys Zemax OpticStudio to about 0.02 RMSE, and that convolving these PSFs onto fisheye automotive simulations adds realistic spatially…","keywords":["automotive simulation","fisheye camera","point spread function","ray tracing","KrakenOS","optical aberration","MTF50","image sharpness"],"falsifier":"Compute KrakenOS and Zemax point spread functions for the same doublet at larger field angles, such as 10, 30, and 60 degrees, and compare them; if the RMSE grows well beyond 0.02, or if a fisheye image convolved with wide-angle PSFs no longer shows the reported center-versus-periphery sharpness pattern, then the equivalence claim holds only for the narrow field of view tested.","tokens_in":6833,"feed_emoji":"🚗","tokens_out":5740,"duration_ms":50926,"temperature":0.7,"pith_summary":"This paper tries to establish that an open-source Python ray-tracing library, KrakenOS, can replace a commercial optical design tool when adding lens artifacts to automotive simulation imagery. It builds the same optical doublet in KrakenOS and in Ansys Zemax OpticStudio and reports that the generated point spread functions agree to about 0.02 root mean square error across red, green, and blue wavelengths. It then convolves those point spread functions onto fisheye simulation images from SynWoodscape and Parallel Domain Woodscape, producing spatially varying blur that is strongest toward the image periphery. The reason this matters is that synthetic training data for autonomous driving can then include realistic lens degradation without needing expensive industrial software.","feed_headline":"Open-source ray tracing matches industrial optics tool at 0.02 RMSE","feed_subtitle":"Convolving the simulated PSFs onto fisheye driving scenes adds realistic, spatially varying blur.","key_machinery":"The load-bearing object is an optical doublet system under test, specified by six surfaces (object plane, BK7 front element, F2 rear element, air gap, aperture stop, and image plane). Around this system the paper builds PSF grids: an 8x8 and a 32x32 array of 160x160-pixel point spread functions, computed separately for red (656 nm), green (531 nm), and blue (486 nm) light over a ±2 degree field of view. Each grid tile is convolved with the spatially corresponding region of a 1280x960 fisheye image, so the PSF grid is the mechanism that turns ray-tracing output into image degradation. Matching these PSFs to Zemax is what supports the claim of industrial equivalence, and convolving them is what carries the optical artifacts into the simulations.","core_discovery":"On the paper's own terms, the central discovery is that KrakenOS-generated Fraunhofer point spread functions for a BK7-F2 optical doublet are nearly indistinguishable from Ansys Zemax OpticStudio's, with RMSE values around 0.02 for most wavelength and field-angle combinations and the green channel at 531 nm matching best. When arranged into 8x8 or 32x32 RGB PSF grids and convolved onto two fisheye automotive simulations, these PSFs transfer the doublet's coma and astigmatism into the images, blurring the periphery more than the center. Measured with the ISO 12233:2023 slanted-edge method, MTF50 drops by about 0.057 cycles/pixel in the middle region and 0.086 cycles/pixel at the edge, while the center is essentially unchanged. The authors conclude that KrakenOS can match industrial performance and that the degradation is spatially variant rather than a uniform Gaussian blur.","pith_inferences":["The authors' own caveat implies the pipeline becomes realistic only when the simulation warp and the convolution use the same lens; a natural extension is to estimate or measure the actual Woodscape fisheye lens PSFs and degrade with those.","The ±2 degree field of view of the doublet is far smaller than a fisheye's field of view, so the reported spatially varying pattern should be treated as a demonstration of the mechanism rather than a model of a real fisheye lens.","A testable extension is to feed both baseline and degraded simulations to a monocular depth estimator; given prior evidence that chromatic aberration and field-dependent blur affect depth accuracy, the degraded set should reproduce more of that behavior.","Porting KrakenOS's ray tracing to a GPU would make fine PSF grids practical at full fisheye resolution, turning this offline pipeline into a plausible real-time augmentation for large-scale training."],"forward_implications":["Synthetic automotive data can be optically degraded with an open-source ray tracer, lowering the cost of lens-realistic simulation.","Because the blur is spatially varying, downstream perception models trained on degraded images may learn to cope with field-dependent sharpness and aberrations rather than uniform Gaussian blur.","The same pipeline can be reused for more complex lenses by swapping the doublet for a different system under test and recomputing its PSF grid.","Sharpness measured by MTF50 gives a quantitative way to compare degraded simulations with real camera behavior, enabling lens-aware sim-to-real transfer."],"supporting_citations":[{"why":"Supplies the open-source KrakenOS ray-tracing library that generates the PSF grids under test.","marker":"[8]"},{"why":"Industrial benchmark Ansys Zemax OpticStudio whose PSFs are the comparison target.","marker":"[9]"},{"why":"Provides the SynWoodscape fisheye simulation image that the pipeline degrades.","marker":"[10]"},{"why":"Provides the Parallel Domain Woodscape fisheye simulation image that the pipeline degrades.","marker":"[11]"},{"why":"ISO 12233:2023 standard that defines the slanted-edge method used to measure sharpness.","marker":"[12]"},{"why":"Demonstrates that optical artifacts are spatially variant, motivating the PSF-grid degradation approach.","marker":"[6]"},{"why":"Supplies the synthetic-scenes spatial frequency response method used to compute MTF50 values.","marker":"[17]"},{"why":"Defines the radial annuli into which slanted-edge measurements are grouped for center, middle, and edge regions.","marker":"[7]"}],"fun_headline_variants":["Open-source ray tracing matches Zemax at 0.02 RMSE","KrakenOS matches industrial optics for fisheye sim","Realistic optical aberrations added to synthetic fisheye driving scenes","Open-source optics sim brings spatially varying blur to automotive scenes","Open-source ray tracing RMSE 0.02 vs Zemax for driving sims"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that a simple doublet's point spread functions, computed over a ±2 degree field of view, represent the optical blur of the fisheye camera whose simulation images are being degraded.","fun_headline_variants_meta":{"raw":{"variants":["Open-source ray tracing matches Zemax at 0.02 RMSE","KrakenOS matches industrial optics for fisheye sim","Realistic optical aberrations added to synthetic fisheye driving scenes","Open-source optics sim brings spatially varying blur to automotive scenes","Open-source ray tracing RMSE 0.02 vs Zemax for driving sims"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001295,"raw_usage":{"total_tokens":5288,"prompt_tokens":947,"completion_tokens":4341,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":563,"completion_tokens_details":{"reasoning_tokens":4248}},"tokens_in":563,"tokens_out":4341,"duration_ms":30159,"temperature":1.0,"reasoning_tokens":4248,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T19:48:38.469354+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute KrakenOS and Zemax point spread functions for the same doublet at larger field angles, such as 10, 30, and 60 degrees, and compare them; if the RMSE grows well beyond 0.02, or if a fisheye image convolved with wide-angle PSFs no longer shows the reported center-versus-periphery sharpness pattern, then the equivalence claim holds only for the narrow field of view tested.","supporting_citations":[{"cited_title":"Krakenos: Python-based general exact ray trac- ing library,","cited_arxiv_id":null,"evidence_quote":"Supplies the open-source KrakenOS ray-tracing library that generates the PSF grids under test."},{"cited_title":"Ansys zemax opticstudio comprehensive optical design software,","cited_arxiv_id":null,"evidence_quote":"Industrial benchmark Ansys Zemax OpticStudio whose PSFs are the comparison target."},{"cited_title":"Synwoodscape: Synthetic surround-view fisheye camera dataset for autonomous driving,","cited_arxiv_id":null,"evidence_quote":"Provides the SynWoodscape fisheye simulation image that the pipeline degrades."},{"cited_title":"Parallel domain synthetic data for the woodscape challenge 2023,","cited_arxiv_id":null,"evidence_quote":"Provides the Parallel Domain Woodscape fisheye simulation image that the pipeline degrades."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"ISO 12233:2023 standard that defines the slanted-edge method used to measure sharpness."},{"cited_title":"High-quality computational imaging through simple lenses,","cited_arxiv_id":null,"evidence_quote":"Demonstrates that optical artifacts are spatially variant, motivating the PSF-grid degradation approach."},{"cited_title":"Ss-sfr: Synthetic scenes spatial frequency response on virtual kitti and degraded au- tomotive simulations for object detection,","cited_arxiv_id":null,"evidence_quote":"Supplies the synthetic-scenes spatial frequency response method used to compute MTF50 values."},{"cited_title":"Measuring natural scenes sfr of automotive fisheye cameras,","cited_arxiv_id":null,"evidence_quote":"Defines the radial annuli into which slanted-edge measurements are grouped for center, middle, and edge regions."}],"review_version":1}